Evaluation of the Lumipulse G TP-N Chemiluminescent Immunoassay as a Syphilis Screening Test

KEYWORDS: syphilis, treponemal test, Lumipulse G TP-N, chemiluminescent enzyme immunoassay

INTRODUCTION

Syphilis is a sexually transmitted disease caused by the bacterium Treponema pallidum subspecies pallidum (T. pallidum). The disease progresses through four stages: primary, secondary, latent, and tertiary (late). The primary stage presents as a chancre at the initial site of infection, which is followed by systemic secondary-stage symptoms of headache, fatigue, and body rash (typically involving the palms of the hands and soles of the feet). If left untreated, the infection can enter a latent stage where symptoms recede, leaving no signs of syphilis infection. The latent period can last for years until symptoms return during tertiary syphilis which can affect multiple organ systems and lead to serious complications, including death (1).

Although treatment for syphilis is readily available, syphilis rates continue to rise in the United States. In 2015, the rate of reported syphilis reached an all-time high of 23.6 cases per 100,000 population, representing a more than 2-fold increase in the past 10 years (1). Most of these cases are attributed to men, particularly men who have sex with men (MSM). However, from 2014 to 2015, increased rates were also seen in heterosexual men and women and in subjects with congenital infections (2). To prevent syphilis transmission, routine syphilis screening is now recommended for high-risk individuals (e.g., HIV-infected individuals, incarcerated individuals, and MSM) and pregnant woman (3,–6).

A syphilis screen is commonly performed using a serological assay to detect either nontreponemal antibodies directed against lipoidal antigens released from treponemes and damaged host cells or treponemal antibodies directed against T. pallidum proteins. There are currently two approaches to syphilis screening: the traditional screening algorithm and the reverse-screening algorithm. The traditional screening algorithm uses a nontreponemal test (e.g., rapid plasma reagin [RPR]) as the initial screening method. If the specimen is reactive, a confirmatory treponemal test is performed. The reverse-screening algorithm uses a treponemal test for screening, followed by a nontreponemal test if the initial screen is reactive. A reactive nontreponemal test indicates active syphilis. If a discrepancy exists between the treponemal screen and the nontreponemal test, a second treponemal assay (e.g., T. pallidum particle agglutination [TP·PA]) is performed to determine a true syphilis infection (usually representing a past treated infection) or a false-positive treponemal screening test. Many high-volume laboratories have adopted the reverse-screening algorithm due to the availability of automated treponemal antibody detection systems, which can produce faster screening results at lower costs (7, 8).

The Lumipulse G TP-N test is a fully automated, chemiluminescent enzyme immunoassay that qualitatively detects IgG and IgM antibodies against T. pallidum antigens in human serum and plasma. The performance characteristics and workflow efficiency of the Lumipulse G TP-N were compared to those of the Bioplex 2200 Syphilis IgG multiplex flow immunoassay, another automated treponemal assay (9, 10). Results that were reactive by either screening assay were followed up with supplemental RPR and T. pallidum particle agglutination (TP·PA) tests. The goal of this study was to evaluate the utility of Lumipulse G TP-N as an initial screening test to aid in the diagnosis of syphilis.

Go to:

RESULTS

Two-method analysis of routine samples.

A total of 4,034 serum samples routinely submitted for syphilis screening were tested using the Bioplex 2200 Syphilis IgG and Lumipulse G TP-N assays. Percent agreement between the two automated tests was excellent at 99.0% (κ = 0.88; 95% confidence interval [CI], 98.7% to 99.3%) (Table 1). Initial syphilis screening of the 4,034 specimens yielded a total of 202 (5.0%) reactive samples by the Bioplex 2200 Syphilis IgG and Lumipulse G TP-N assays. All reactive screening specimens were further tested by TP·PA and RPR supplemental tests. Of the 202 samples with a reactive syphilis screen result, 78.7% (159) were TP·PA positive, 1.0% (2) were TP·PA inconclusive, and 32.7% (66) had RPR titers (range, 1:1 to 1:128). Using the reverse-screening algorithm interpretation, the Bioplex 2200 Syphilis IgG and Lumipulse G TP-N screening assays would have identified 159 and 157 syphilis infections, respectively.

TABLE 1

Comparison of Bioplex 2200 Syphilis IgG and Lumipulse G TP-N assay results^a

Lumipulse G TP-N result	No. (%) of Bioplex 2200 Syphilis IgG results
	Positive	Negative	Total
Positive	161 (82.6)	7 (0.2)	168 (4.2)
Negative	34 (17.4)	3,832 (99.8)	3,866 (95.8)
Total	195 (4.8)	3,839 (95.2)	4,034

^aEquivocal results were counted as positive. Agreement, 99.0% (95% CI, 98.7 to 99.3); κ value, 0.88.

Samples with RPR titers (n = 66) displayed complete agreement with the Bioplex 2200 Syphilis IgG and Lumipulse G TP-N assay results, while the TP·PA test displayed some discrepancies. In comparisons between the TP·PA test and the Lumipulse G TP-N assay, the positive percent agreement (PPA) and negative percent agreement (NPA) were 98.7% and 78.0%, respectively (Table 2). The Bioplex 2200 Syphilis IgG assay displayed a similar positive percent agreement (100%) but poor negative percent agreement (17.1%). Overall, a total of 34 Bioplex 2200 Syphilis IgG results and 11 Lumipulse G TP-N results did not correlate with the TP·PA assay results. Patient chart reviews of these discrepant results did not reveal any evidence of syphilis infection, indicating that the Lumipulse G TP-N screening assay generated 25 fewer falsely reactive results.

TABLE 2

Comparison of Bioplex 2200 Syphilis IgG and Lumipulse G TP-N test results to TP·PA test results during routine syphilis screening^a

Assay result	No. of TP·PA results (n = 200)		PPA (95% CI)	NPA (95% CI)	PA (95% CI)	κ value
	Positive	Negative
Bioplex 2200 Syphilis IgGb
Positive	159	34	100 (97–100)	17.1 (8.2–31.6)	83 (77.2–87.6)	0.25
Negative	0	7
Lumipulse G TP-N
Positive	157	9	98.7 (95.2–99.9)	78.0 (63.1–88.2)	94.5 (90.3–97)	0.82
Negative	2	32

^aPPA, positive percent agreement; NPA, negative percent agreement; PA, percent agreement.

^bData include equivocal results.

Two-method analysis of HIV-positive samples.

A retrospective analysis was performed on 100 HIV-infected patients to assess the performance of the Lumipulse G TP-N screening assay in a high-risk patient population. Similarly to the routine sample analysis, the agreement between the Bioplex 2200 Syphilis IgG and Lumipulse G TP-N assays was excellent at 98% (κ, 0.95; 95% CI, 94.2% to 100%). A total of 29 of the 100 HIV-infected samples were reactive by either automated screening test, with 24 TP·PA-positive results, 2 inconclusive TP·PA results, and 16 with RPR titers (range, 1:2 to 1:64). Discordances were seen between the results from the two automated assays and the TP·PA results (Table 3). Upon patient chart review, fewer (n = 2) presumptive falsely reactive results were produced by the Lumipulse G TP-N assay.

TABLE 3

Comparison of Bioplex 2200 Syphilis IgG and Lumipulse G TP-N within an HIV positive population^a

Assay result	No. of TP·PA results (n = 27)		PPA	NPA	PA
	Positive	Negative
Bioplex 2200b
Positive	24	3	100	0	88.9
Negative	0	0
Lumipulse G TP-N
Positive	24	1	100	66.7	96.3
Negative	0	2

^aPPA, positive percent agreement; NPA, negative percent agreement; PA, percent agreement.

^bData include equivocal results.

Workflow efficiency.

The workflow efficiency of each automated testing system was analyzed by timing each stage of the syphilis screening process. Daily instrument and reagent preparations prior to patient testing were similar for all of the instrument systems. Once patient specimens were loaded onto the instrument, the Lumipulse G1200 system produced shorter times to first results (15 min less) and to subsequent results (5 s less) than the Bioplex 2200 IgG system. As a result, the assay run time for 100 syphilis screens was reduced by 25 min using the Lumipulse G1200 system. Although shutdown and start-up times were recorded for both instruments, only the Lumipulse G1200 system was safe to shut down, since the computer system can be turned off without losing power to the refrigeration unit that maintains reagent stability.

Go to:

DISCUSSION

The absence of a gold standard for the immunologic detection of syphilis infection has made syphilis diagnosis challenging for clinical laboratories. The United States Centers for Disease Control and Prevention (CDC) endorse the traditional RPR-based screening algorithm, while the Association of Public Health Laboratories, the United Kingdom Health Protection Agency, and the International Union against Sexually Transmitted Infections support the reverse-screening algorithm (8, 11,–13). These conflicting recommendations stem from the differing performance characteristics observed with each screening assay. Nontreponemal assays generate fewer falsely reactive results, while treponemal tests are considered to be more sensitive, particularly at detecting early and latent stages of syphilis (8, 14,–16). Treponemal screening assays also have the added benefit of being automated, which can improve workflow and reduce costs in high-volume laboratories.

The automated Lumipulse G TP-N assay was cleared in July 2016 by the U.S. Food and Drug Administration (FDA) as an aid in the diagnosis of syphilis. To our knowledge, this is the only study comparing the Lumipulse G TP-N assay to another automated, FDA-cleared assay (17,–19). A total of 4,034 serum samples routinely submitted for syphilis screening and samples from 100 high-risk HIV patients were tested by the Lumipulse G TP-N assay and the Bioplex 2200 Syphilis IgG assay. The Lumipulse G TP-N assay showed excellent agreement with the Bioplex 2200 Syphilis IgG assay in routine (99.0%) and HIV-positive (98.0%) samples. Collectively, only 43 discrepant results were seen between the two automated assays.

Additional testing to determine if these discrepant results represent truly reactive results or falsely reactive results from the two assay systems was not performed. However, 95% (41/43) of these discrepancies likely represent falsely reactive results given that the TPPA assays were nonreactive, only one of automated treponemal assays produced a reactive result, and patient chart reviews did not reveal any indication of syphilis infection. The Bioplex 2200 Syphilis IgG assay falsely reactive results (n = 34) ranged from weakly reactive (0.9) to strongly reactive (8.0). Of note, only 1 of 6 equivocal Bioplex 2200 Syphilis IgG results was considered a truly reactive result. On the other hand, the Lumipulse G TP-N test produced only 7 falsely reactive results, all of which were weakly reactive, with cutoff index (C.O.I.) values between 1.2 to 2.5.

Overall, the Lumipulse G TP-N test produced 27 fewer falsely reactive results in both patient cohorts. The discrepancy in results may be attributed to slight differences between the two automated assays. The Lumipulse G TP-N test uses two recombinant antigens (Tp15-17 and TpN47), while the Bioplex 2200 IgG assay uses three recombinant T. pallidum proteins (15 kDa, 17 kDa, and 47 kDa). The recombinant Tp15-17 antigens used in the Lumipulse G TP-N assay may have masked epitopes that caused falsely reactive Bioplex 2200 Syphilis IgG results. Also, each assay uses a different detection method with different reagents and beads that may contribute to the discrepant results. The lower level of falsely reactive results produced by the Lumipulse G TP-N assay indicates that using this assay for syphilis screening can reduce the amount of additional confirmatory RPR and TP·PA testing in a reverse syphilis screening algorithm.

However, there were two samples among the discrepant results that the Lumipulse G TP-N assay may have incorrectly interpreted as nonreactive (cutoff index values of 0.1 and 0.7). In these two samples, the Lumipulse G TP-N and RPR assays were nonreactive, while the Bioplex 2200 Syphilis IgG and the TP·PA assays were reactive. According to the reverse-screening algorithm, these results represent either a past or present syphilis infection. Even though neither patient had any signs or history of syphilis infection, it is possible that primary syphilis symptoms may have been missed, and a new patient sample should be submitted 2 to 3 weeks later for follow-up testing to confirm these results.

This evaluation was limited by not having a true gold standard for serologic syphilis testing. Without a gold standard, discrepant results can be difficult to interpret, but correlating the results between four different assays and patient chart reviews can support a falsely reactive or nonreactive result. Another limitation was the lack of symptomatic patients to determine the stage of disease. Chart reviews of the 202 reactive syphilis screens revealed only 7 patients with syphilis-like symptoms, which included rash, headache, joint pain, and fatigue. The inability to determine the stage of disease prevented the assessment of each assay at various stages of syphilis infection.

Overall, the Lumipulse G TP-N assay offers improved workflow compared to the Bioplex 2200 Syphilis IgG assay and a lower rate of false reactivity. In laboratories performing high-volume syphilis testing, the automated Lumipulse G TP-N assay represents a suitable high-throughput option for initial syphilis screening.

Go to:

MATERIALS AND METHODS

Study cohorts.

A prospective analysis of 4,034 serum specimens routinely submitted for syphilis screening from September and October of 2017 and a retrospective analysis of 100 samples from HIV-infected patients were performed by both automated systems. Specimens submitted for routine syphilis screening included samples from obstetrics/gynecology (OB/GYN) departments, labor and delivery departments, incarcerated persons, and miscellaneous clinic and hospitalized patients.

Ethics statement.

The University of Texas Medical Branch Institutional Review Board determined that this activity represented quality improvement (assay validation) and was therefore exempt from Institutional Review Board (IRB) review.

BioPlex 2200 Syphilis IgG assay.

Serum specimens submitted for routine syphilis screening were tested on a Bioplex 2200 analyzer according to the manufacturer's instructions. The BioPlex 2200 Syphilis IgG assay (Bio-Rad Laboratories, Hercules, CA) is a multiplex flow immunoassay (MFI) that qualitatively detects IgG antibodies against Treponema pallidum in human serum and is the current assay used by the University of Texas Medical Branch clinical laboratory for syphilis testing.

The MFI method can detect multiple antigen-specific antibodies using dyed, antigen-coated beads. In the BioPlex 2200 Syphilis IgG assay, three different sets of dyed beads coated with recombinant T. pallidum proteins (15 kDa, 17 kDa, and 47 kDa) are incubated with patient serum. After a washing step, anti-human IgG conjugated to phycoerythrin, a fluorescent protein, is added to the beads in the reaction chamber. These beads are then washed and passed through a detector. The antigen-coated beads are identified by the fluorescence of the dye, and the amount of IgG captured is determined by the intensity of the phycoerythrin fluorescence. Results are reported as an antibody index (AI) with a range of <0.2 to >8.0 and interpreted as nonreactive (≤0.8), reactive (≥1.1), or equivocal (0.9 or 1.0). For this study, all equivocal results (n = 6) were interpreted as reactive.

Lumipulse G TP-N assay.

Routine serum specimens were collected following Bioplex 2200 Syphilis IgG analysis and were tested the same day on a Lumipulse G1200 analyzer according to the manufacturer's instructions. Frozen serum samples from HIV-infected patients were thawed to room temperature prior to screening on the Lumipulse G1200 analyzer. The Lumipulse G TP-N assay (Fujirebio, Inc., Tokyo, Japan) is a chemiluminescent immunoassay that qualitatively detects IgG and IgM antibodies against Treponema pallidum antigens in human serum and plasma.

The Lumipulse G TP-N chemiluminescent immunoassay utilizes ferrite microparticles coated with recombinant T. pallidum antigens (Tp15-17 and TpN47) to detect anti-T. pallidum antibodies through the formation of immunocomplexes. After a washing step to remove unbound material, T. pallidum antigens labeled with alkaline phosphatase are added to form additional complexes. Following another wash cycle, a substrate solution is added, which triggers a chemiluminescent reaction in the presence of alkaline phosphatase. The light emitted is used to calculate the level of anti-T. pallidum antibodies present based on the calibration data. Results are reported as cutoff index (C.O.I.) values with a range of 0.1 to 100 and interpreted as reactive (≥1.0) or nonreactive (<1.0).

Reflex testing.

Specimens with reactive and equivocal results from BioPlex 2200 Syphilis IgG and Lumipulse G TP-N screening assays were further tested by TP·PA and RPR tests. Commercial kits for TP·PA (Serodia; Fujirebio, Inc., Tokyo, Japan) and RPR (Sure-Vue; Biokit USA, Inc., Lexington, MA) testing were used according to the instructions of the manufacturers. TP·PA test results were read manually at a final serum dilution of 1:80. RPR test results were read manually on 2-fold serial dilutions of serum ranging from 1:1 to 1:512.

Workflow efficiency.

A digital stopwatch was used to record the time required to perform daily and monthly maintenance, instrument start-up and shutdown, and assay preparation. Each task was performed by an experienced technician familiar with each assay's operating system. Testing on both automated systems was performed according to the instructions of the manufacturers. Times were recorded in duplicate, and averages of the data determined at the two times were reported.

Data analysis.

Percentages of agreement and kappa coefficients were calculated to compare the qualitative results from the Lumipulse G TP-N test and the BioPlex 2200 Syphilis IgG assay. Confidence intervals of proportions were calculated using the GraphPad QuickCalcs website.

Go to:

ACKNOWLEDGMENTS

We thank the laboratory technologists of the University of Texas Medical Branch Microbiology Laboratory for assisting with this study and Fujirebio for providing the kits and reagents used during the study.

Go to:

REFERENCES

1. Hook EW., III 16 December 2016. Syphilis. Lancet doi:10.1016/s0140-6736(16)32411-4. [CrossRef] [Google Scholar]
2. CDC. 2016. Sexually transmitted disease surveillance 2015. U.S. Department of Health and Human Services, Atlanta, GA. [Google Scholar]
3. Gray RT, Hoare A, Prestage GP, Donovan B, Kaldor JM, Wilson DP. 2010. Frequent testing of highly sexually active gay men is required to control syphilis. Sex Transm Dis 37:298–305. [PubMed] [Google Scholar]
4. Lewis FM, Schillinger JA, Taylor M, Brewer TH, Blank S, Mickey T, Furness BW, Anschuetz GL, Salmon ME, Peterman TA. 2011. Needle in a haystack: the yield of syphilis outreach screening at 5 US sites—2000 to 2007. J Public Health Manag Pract 17:513–521. doi:10.1097/PHH.0b013e3182113954. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
5. Workowski KA, Bolan GA; Centers for Disease Control and Prevention . 2015. Sexually transmitted diseases treatment guidelines, 2015. MMWR Recomm Rep 64(RR-03):1–137. [PMC free article] [PubMed] [Google Scholar]
6. Branger J, van der Meer JT, van Ketel RJ, Jurriaans S, Prins JM. 2009. High incidence of asymptomatic syphilis in HIV-infected MSM justifies routine screening. Sex Transm Dis 36:84–85. doi:10.1097/OLQ.0b013e318186debb. [PubMed] [CrossRef] [Google Scholar]
7. CDC. 2008. Syphilis testing algorithms using treponemal tests for initial screening—four laboratories, New York City, 2005–2006. MMWR Morb Mortal Wkly Rep 57:872–875. [PubMed] [Google Scholar]
8. Loeffelholz MJ, Binnicker MJ. 2012. It is time to use treponema-specific antibody screening tests for diagnosis of syphilis. J Clin Microbiol 50:2–6. doi:10.1128/JCM.06347-11. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
9. Gomez E, Jespersen DJ, Harring JA, Binnicker MJ. 2010. Evaluation of the Bio-Rad BioPlex 2200 syphilis multiplex flow immunoassay for the detection of IgM- and IgG-class antitreponemal antibodies. Clin Vaccine Immunol 17:966–968. doi:10.1128/CVI.00086-10. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
10. Marangoni A, Nardini P, Foschi C, Moroni A, D'Antuono A, Bacchi Reggiani L, Cevenini R. 2013. Evaluation of the BioPlex 2200 syphilis system as a first-line method of reverse-sequence screening for syphilis diagnosis. Clin Vaccine Immunol 20:1084–1088. doi:10.1128/CVI.00316-13. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
11. Egglestone SI, Turner AJ. 2000. Serological diagnosis of syphilis. PHLS Syphilis Serology Working Group. Commun Dis Public Health 3:158–162. [PubMed] [Google Scholar]
12. French P, Gomberg M, Janier M, Schmidt B, van Voorst Vader P, Young H. 2009. IUSTI: 2008 European Guidelines on the Management of Syphilis. Int J STD AIDS 20:300–309. doi:10.1258/ijsa.2008.008510. [PubMed] [CrossRef] [Google Scholar]
13. CDC. 2011. Discordant results from reverse sequence syphilis screening—five laboratories, United States, 2006–2010. MMWR Morb Mortal Wkly Rep 60:133–137. [PubMed] [Google Scholar]
14. Binnicker MJ, Jespersen DJ, Rollins LO. 2012. Direct comparison of the traditional and reverse syphilis screening algorithms in a population with a low prevalence of syphilis. J Clin Microbiol 50:148–150. doi:10.1128/JCM.05636-11. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
15. Mishra S, Boily MC, Ng V, Gold WL, Okura T, Shaw M, Mazzulli T, Fisman DN. 2011. The laboratory impact of changing syphilis screening from the rapid-plasma reagin to a treponemal enzyme immunoassay: a case-study from the Greater Toronto Area. Sex Transm Dis 38:190–196. doi:10.1097/OLQ.0b013e3181f07e91. [PubMed] [CrossRef] [Google Scholar]
16. Gratrix J, Plitt S, Lee BE, Ferron L, Anderson B, Verity B, Prasad E, Bunyan R, Zahariadis G, Singh AE. 2012. Impact of reverse sequence syphilis screening on new diagnoses of late latent syphilis in Edmonton, Canada. Sex Transm Dis 39:528–530. doi:10.1097/OLQ.0b013e31824e53f7. [PubMed] [CrossRef] [Google Scholar]
17. Li L, Cai B, Tao C, Wang L. 2016. Performance evaluation of CLIA for Treponema pallidumspecific antibodies detection in comparison with ELISA. J Clin Lab Anal 30:216–222. doi:10.1002/jcla.21839. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
18. An J, Chen Q, Liu Q, Rao C, Li D, Wang T, Tao C, Wang L. 2015. Evaluation of the HISCL anti-Treponema pallidumassay as a screening test for syphilis. Clin Vaccine Immunol 22:817–822. doi:10.1128/CVI.00116-15. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
19. Chen Q, An J, Rao C, Wang T, Li D, Feng S, Tao C. 2016. Performance evaluation of a novel chemiluminescence assay detecting Treponema pallidumantibody as a syphilis screening method. Clin Lab 62:519–526. [PubMed] [Google Scholar]

Articles from Journal of Clinical Microbiology are provided here courtesy of American Society for Microbiology (ASM)